Written by Scott Rowley
Just a starter page that will grow as I find useful awk commands to share.
Convert Linux .txt to Windows .txt
Replace the end of a line with a carriage. Useful for converting .txt files from Linux to Windows.
awk 'sub("$", "\r")' linuxfile.txt > windowsfile.txt
Convert Windows .txt to Linux .txt
And the reverse. Useful for converting .txt files from Windows to Linux.
awk '{ sub("\r$", ""); print }' windowsfile.txt> linuxfile.txt
Written by Scott Rowley
In order to remove tabs from your output (say to get a specific variable) use the ‘tr’ command, for example
In my zone.file the line containing “Serial” looks like the following:
If I performed
grep Serial zone.file
I would get the following output
2009110904 ; Serial (yyyymmdd##)
You can’t see them here but there are several tabs before that date. In my example I want to get just the actual serial number (2009110904). To do this I’ll first strip off the “; Serial…” information;
grep Serial zone.file | awk -F; {'print $1'}
This command is telling it to look for the ; special character and break up the output there, then from there it is told to print the first column of information this results in:
2009110904
The problem here is that if we want this to be our variable we are actually working with ” 2009110904″ instead of the desired “2009110904” In this particular case we know those are tabs not spaces. So we’ll need to use tr in order to remove those tabs:
grep Serial zone.file | awk -F\; {'print $1'} | tr -d 't'
This will return to us exactly what we want
2009110904
Written by Scott Rowley
Recently I had need of cutting up domain names such as domain.com, this is easy enough for as simple a domain as this. However, the problem arises when you need to cut up something more like “education.k12.ia.us”. In order to get the extension alone in either case you can do the following:
echo domain.com | awk -F. {'print $2'}
com
Now normally if you have a domain as simple as example.com with only one period then you can can just do the opposite of this and print $1:
echo domain.com | awk -F. {'print $1'}
domain
The problem with more complex URLs is that awk’s field delimiter is now looking at multiple periods to look at, could be just one or it could be two, three, four or more.
To get the final extension in this case then we’ll use the following:
echo education.k12.ia.us | awk -F. {'print $NF'}
us
Then for the rest of it we’ll have to reverse it, then cut off the first field then reverse it again:
echo education.k12.ia.us | rev | cut -d. -f2- | rev education.k12.ia